Preprints (www.preprints.org) | NOT PEER-REVIEWED | Posted: 8 October 2018 doi:10.20944/preprints201808.0496.v2 Hershkovitz Montiaceae Synopsis of a new taxonomic synthesis of Montiaceae (Portulacineae) based on rational metadata analysis, with critical new insights on historically poorly understood taxa and a review of ecological evolution and phylogeography Mark Alan HERSHKOVITZ1 1Santiago, Chile [email protected] Abstract: Montiaceae (Portulacineae) comprise a clade of at least 280 species and ca. 30 subspecific taxa primarily of western America and Australia. This work uses existing phylogenetic metadata to elaborate a new cladistic taxonomic synthesis, and clarifies morphological circumscriptions of several poorly known species. A total of 20 taxa are validated, seven new and 13 necessary nomenclatural recombinations. Hypotheses of Montiaceae historical biogeography and phenotypic evolution are evaluated in light of recent metadata. Key words: Montiaceae, taxonomy, phylogeny, ecology, phylogeography, evolution. 1. Introduction This work presents a new cladistic taxonomy of Montiaceae (Portulacineae) and several of its included taxa, along with notes on the diagnostics of certainly poorly known species and a summary of new interpretations of phylogeography and phenotypic and ecological evolution. The present work includes 20 nomenclatural novelties. However, the whole of the novelty is greater than the sum of these parts. The generic circumscriptions and diversity estimates are modified from Hernández-Ledesma et al. (2015).The suprageneric taxonomy is the first proposed since McNeill (1974) and the only phylogenetic one. Critical reevaluation of certain common and usually misidentified Chilean taxa is the first since Reiche (1898). Existing metadata are interpreted as evidence for a hybrid origin of a genus. And existing metadata are interpreted as evidence for decoupling of cladogenesis with phenotypic evolution. Montiaceae comprise a clade of at least 280 species plus ca. 30 additional accepted subspecific taxa, primarily of western America and Australia (Hernández-Ledesma et al. 2015; Ogburn & Edwards 2015). The species diversity estimate (calculated from the taxonomic treatment) is almost 25% greater than that reported in Hernández-Ledesma (2015) and reflects more recent and ongoing research. Formerly, Montiaceae were included in the traditional but paraphyletic family Portulacaceae (Nyffeler & Eggli 2010). Its conceptual origins trace to Hershkovitz (1993, who referred to a group he called Western American Portulacaceae. Molecular analysis (Hershkovitz & Zimmer 1997) showed that this group comprised most of a clade of Portulacaceae they called the PAW clade, comprising: Phemeranthus Raf., the Australian genus Rumicastrum Ulbrich, and geographically western American genera of Portulacaceae. The current circumscription of Montiaceae was completed with the addition of Hectorelleae (Applequist et al. 2006; Wagstaff & Hennion 2006). The modern cladistic generic classification originated with morphological analyses of Carolin (1987, 1993) and Hershkovitz (1991a, 1993). Slowly, the new familial and generic concepts began to replace the traditional concepts of Pax & Hoffmann (1934) and McNeill (1974). Molecular evidence, beginning with Hershkovitz & Zimmer (1997), stimulated further acceptance of the cladistic classification and fomented several subsequent studies. Here I 1 © 2018 by the author(s). Distributed under a Creative Commons CC BY license. Preprints (www.preprints.org) | NOT PEER-REVIEWED | Posted: 8 October 2018 doi:10.20944/preprints201808.0496.v2 Hershkovitz Montiaceae elaborate a modified cladistic supraspecific taxonomy based on molecular evidence, especially Hershkovitz (2006), Ogburn & Edwards (2015), and Hancock et al. (2018). This evidence is reviewed in Hershkovitz (2018a). This work first summarizes information on the ecology and phylogeographic history of Montiaceae. The taxonomy follows. The generic taxonomy supersedes that of Hernández- Ledesma et al. (2015), which superseded that of Carolin (1993). In addition, I comment on the taxonomy and taxonomic history of selected taxa, especially Calandrinia Kunth, Rumicastrum, and Cistanthe Spach and recombine binomials into Cistanthe, Calyptridium Nuttall and Philippiamra Kuntze. I conclude that Lewisiopsis has a hybrid origin and represents a living fossil. Finally, I offer detailed observations on taxonomically troublesome Chilean taxa. 2. Montiaceae ecology and ecological evolution Montiaceae species distinguish from other Portulacineae phylogeographically, ecologically, and phenotypically. Their distribution concentrates in the high-relief and climatically diverse temperate zone westwards from the American high cordillera (cis-cordilleran), with lesser additional native diversity east of the American cordillera (trans-cordilleran), Siberia, Australia, New Zealand, and the Kerguelen Islands (Hershkovitz 2018a, b). Correspondingly, Montiaceae species concentrate especially in cooler/moister (higher altitude and/or latitude) and warmer/drier Mediterranean and desert environments (Ogburn & Edwards 2015; Hershkovitz 2018b). They also differ from other Portulacineae in comprising mostly hemicryptophytic (including acaulescent rosettiform-caudiciform) perennials and therophytes (Hershkovitz 1993, 2018b; Ogburn & Edwards 2015) Other Portulacineae are distributed mainly more towards the tropics and from the American high cordillera eastwards (trans-cordilleran) and Africa/Madagascar, with a few taxa in warmer parts of Australasia/Oceania (Hershkovitz 1993, 2018b; Ogburn & Edwards 2015). They are mostly phanerophytic herbs, shrubs, trees, and vines, and often stem succulents. However, most Portulaca (Portulacaceae) species are monocarpic annuals (Ocampo 2012) and most Anacampserotaceae species are hemicryptophytes, a few acaulescent (Carolin 1987, 1993; Rowley 1994). Montiaceae niches are diverse, including desert to aquatic habitats and arctic/alpine to subtropical habitats and everything in between. This has been thought to be related to the diversity of the local habitats in the main range of the family, far-western North America and South America (Hershkovitz 2006; Ogburn & Edwards 2015). But Hershkovitz (2018b) reconsidered this view, pointing out that some Montiaceae lineages in this range never diversified, while Rumicastrum, the largest genus in the family, is endemic to less macroecologically diverse regions of Australia (see below). The same phenomenon characterizes many lineages o f angiosperms in macroecologically diverse western America: some have diversified and some have not. Similarly, other angiosperm lineages have diversified exceptionally in relatively less macroecologically diverse regions of South Africa and Australia. Thus, local macroecological diversity is neither necessary, nor sufficient, to explain Montiaceae diversity. The criterion of necessity and sufficiency of evolutionary explanation was emphasized in Hershkovitz (2018b). Montiaceae include species that span small to large ranges and rare to common frequencies. Several species, as noted in the taxonomy discussion, are adventive outside of their original native ranges. These species concentrate among the genera Claytonia and Montia, and most are 2 Preprints (www.preprints.org) | NOT PEER-REVIEWED | Posted: 8 October 2018 doi:10.20944/preprints201808.0496.v2 Hershkovitz Montiaceae annuals. Otherwise, there appears to be no particular correlation between species range/frequency and other obvious possible phenotypic or genetic factors. Using various methods of statistical phylogenetic comparative analysis, Ogburn & Edwards (2015) studied ecological evolution of Montiaceae, especially the relation between life history and temperature and moisture niche. They reported an increased rate of phylogenetic diversification of temperature niche and life history relative to other Portulacineae. Montiaceae have incurred ca. 14 life history transitions since the split between Cistantheae and Montioideae, 25-35 million years ago, out of the total ca 16 transitions among all Portulacineae over 50+ million years (Ogburn & Edwards 2015; Hershkovitz 2018b). Ogburn & Edwards (2015) concluded that the ancestral growth form of Montiaceae was acaulescent rosettiform-caudiciform and tap-rooted, with little annual above-ground growth. This growth form characterizes, e.g., certain species of Calandrinia sect. Acaules and Lewisia. This conclusion was more or less in agreement with mor5phological cladistic conclusions (Carolin 1987; Hershkovitz 1993). However, Montiaceae also include rhizomatous and tuberous perennials, as well as perennials with substantial above-ground growth, including pachyform shrubs (Hershkovitz 2018b). Ogburn & Edwards (2015) determined that perennials generally were more cold-adapted than annual, and concluded that the annual life history evolved many times as an adaptation to transition to warmer environments. They also concluded that annuals were absent in cold environments because of the shorter growth season, which did not permit annuals to complete their life cycle. Ogburn & Edwards (2015) attributed phylogenetic life history lability of Montiaceae to their ancestral herbaceous and, in particular, acaulescent rosettiform-caudiciform growth form. Phylogenetic niche breadth lability itself had been ascribed previously to herbaceous more than woody growth forms (Smith & Beaulieu 2009). Ogburn & Edwards (2015) argued that the rosettiform-caudiciform form of Montiaceae permitted